Department of Respiratory and Critical Care Medicine, First People's Hospital of Kashi, Kashi, China.
Department of Clinical Research Center of Infectious Diseases (Pulmonary Tuberculosis), First People's Hospital of Kashi, Kashi, China.
Front Public Health. 2022 May 4;10:881234. doi: 10.3389/fpubh.2022.881234. eCollection 2022.
Based on the respiratory disease big data platform in southern Xinjiang, we established a model that predicted and diagnosed chronic obstructive pulmonary disease, bronchiectasis, pulmonary embolism and pulmonary tuberculosis, and provided assistance for primary physicians.
The method combined convolutional neural network (CNN) and long-short-term memory network (LSTM) for prediction and diagnosis of respiratory diseases. We collected the medical records of inpatients in the respiratory department, including: chief complaint, history of present illness, and chest computed tomography. Pre-processing of clinical records with "jieba" word segmentation module, and the Bidirectional Encoder Representation from Transformers (BERT) model was used to perform word vectorization on the text. The partial and total information of the fused feature set was encoded by convolutional layers, while LSTM layers decoded the encoded information.
The precisions of traditional machine-learning, deep-learning methods and our proposed method were 0.6, 0.81, 0.89, and 1 scores were 0.6, 0.81, 0.88, respectively.
Compared with traditional machine learning and deep-learning methods that our proposed method had a significantly higher performance, and provided precise identification of respiratory disease.
基于南疆呼吸系统疾病大数据平台，建立预测和诊断慢性阻塞性肺疾病、支气管扩张症、肺栓塞和肺结核的模型，为基层医生提供辅助。
该方法结合卷积神经网络（CNN）和长短时记忆网络（LSTM）进行呼吸系统疾病的预测和诊断。我们收集了呼吸科住院患者的病历，包括：主诉、现病史和胸部计算机断层扫描。使用“结巴”分词模块对临床记录进行预处理，使用双向编码器表示从变压器（BERT）模型对文本进行词向量化。融合特征集的部分和整体信息通过卷积层进行编码，而 LSTM 层则对编码信息进行解码。
传统机器学习、深度学习方法和我们提出的方法的精度分别为 0.6、0.81、0.89 和 1 分。
与传统机器学习和深度学习方法相比，我们提出的方法性能显著提高，为呼吸系统疾病的精确识别提供了支持。
